, ,

1
????????????? ? ???????????? ????????

• ????????, ?.?. ????????, 2007-2008

2
??????????????? ???????? ????

• ????????? ???? ??????? ????? ...
• ????? ?????? ??? ?? ???????, ??? ???? ???? ??
  ????? ???????? ????????. ...
• ??????? ??????? ?? ????????, ??? ???? ???? ?
  ???-?? ?????????? ???????, ??? ?
• ???????????? ????????? ??????? ???? ?? ????? ??
  ?????. ...
• ??? ?????????? "?????????" ?????? ???????, ???
  ???? ?? ????? ...
• ?????? -- ??????, ??????? ???? ? ?????????
  ???????? ? ?? ??????
• ???, ??-??????, ? ???? ????????? ? ???? ?? ????,
  ?, ??-??????, ??? ????? ?
• ?????? ????? ????????, ?? ??????? ???? ??????,
  ??? ?????, ? ?? ????? ?? ??????
• ? ??? ??? ??????? ???? ?? ????, ???? ? ?????
  ?????, ...
• ... ??????? ? ???? ? ??? ????????, ?????? ????,
  ????.. ...
• ??, ??????, ??? ??? ?? ???????????, ? ????
  ????. ... ...
• ???????, ???? ???? ?? ??? ???...

3
??????????????? ???????? ????

• ????????????-?,??,???? ??????????????,??,??
  , Class"???_???-?" Class"?????" ??????
  ??????????????????????????,??
  ???????????????????,??,????,?? Class"?-??"
  Class"???????????" ???????????????????,???,?
  ?,????,?? ?????? ???????-?
  ????????????????????????????,??,????,??
  Class"???????". (?????? ??????. ????? ?????
  ??????)

4
??????????????? ???????? ????

• ?????
• ????? ???????
• ??????????????? ????? ?????
• ??????????????? ???????? ???????????? ????????
• ????????????? ?????????????? ??????????????
• ??????????
• ????? ????
• ??? ???????
• ?????? ?????????????
• ????????????? ?????????????? ????????????
• ????????????? ????? nomina feminina

5
??????????????? ????????, GNOME

• This tables
• ltne id"ne2" cat"poss-np" per"per3" num"sing"
• gen"neut" gf"subj" lftype"term"
• onto"concrete" ani"inanimate"
• deix"deix-no" count"undersp-count"
• generic"generic-no"gt
• ltne id"ne3" cat"this-np" per"per3" num"sing"
• gen"neut" gf"gen" lftype"term"
• onto"concrete" ani"inanimate"
• deix"deix-yes" count"count-yes"
• structure"atom"
• generic"generic-no"gt

6
??????????????? ????????, GNOME

• (allow)
• ltne id"ne4" cat"bare-np" per"per3" num"plur"
• gen"neut" gf"obj" lftype"term" onto"person"
• ani"animate" deix"deix-no" count"count-yes"
• structure"set" generic"generic-yes"gt
• scholars lt/negt
• (to link)
• ltne id"ne5" cat"pers-pro" per"per3" num"sing"
• gen"neut" gf"obj" lftype"term" onto"concrete"
• ani"inanimate" deix"deix-yes" count"count-yes"
• structure"atom" generic"generic-no"gt it lt/negt

7
????????????? ???????? ?????????

• And 00000000 the 00000000
• soldiers 23241000 platted 21072000
• a 00000000 crown 21110400
• of 00000000 thorns 13010000
• and 00000000 put 21072000
• it 00000000 on 00000000
• his 00000000 head 21030000
• and 00000000 they 00000000
• put 21072000 on 00000000
• him 00000000 a 00000000
• purple 31241100 robe 21110321

8
????????????? ???????? ?????????

• 00000000 Low content word (and, the, a, of, on,
  his, they etc)
• 13010000 Plant life in general
• 21030000 Body and body parts
• 21072000 Object-oriented physical activity (e.g.
  put)
• 21110321 Men's clothing outer clothing
• 21110400 Headgear
• 23231000 War and conflict general
• 31241100 Colour

9
????????????? ???????? Penn Tree Bank

• Vandenberg and Rayburn were wise enough TRACE
  to leave specific operations to presidents.
• baseleave2 tenseinfinitival
• arg2presidents
• arg1specific operations
• arg0TRACE -gt Vandenberg and Rayburn

10
????????????? ???????? Penn Tree Bank

• HIT (sense strike)
• Arg0 hitter
• Arg1 thing hit
• Arg2 instrument, hit with
• HAIL (sense pellets of ice from the sky)
• Labels allow to capture transitivity
  alternations
• John (Arg0) broke the window
• (Arg1) and The window (Arg1) broke.

11
????????????? ???????? Penn Tree Bank

• EDGE (sense move slightly)
• Arg0 causer of motion3
• Arg1 thing in motion
• Arg2 distance moved
• Arg3 start point
• Arg4 end point
• Arg5 direction
• The publishing unit reported revenue edged up
  2.6 to 263.2
• million from 256.6 million.

12
????????????? ???????? Penn Tree Bank

• BUY
• Arg0 buyer
• Arg1 thing bought
• Arg2 seller, bought-from
• Arg3 price paid
• Arg4 benefactive, bought-for

13
????????????? ???????? Penn Tree Bank

• PURCHASE BUY SELL
• Arg0 buyer Arg0 buyer Arg0 seller
• Arg1 thing bought Arg1 thing bought Arg1
  thing sold
• Arg2 seller Arg2 seller Arg2 buyer
• Arg3 price paid Arg3 price paid Arg3 price
  paid
• Arg4 benefactive Arg4 benefactive Arg4
  benefactive

14
????????????? ???????? Penn Tree Bank

• The company bought a wheel-loader from Dresser.
• Arg0 The company
• rel bought
• Arg1 a wheel-loader
• Arg2-from Dresser
• TV stations bought "Cosby" reruns for record
  prices.
• Arg0 TV stations
• rel bought
• Arg1 "Cosby" reruns
• Arg3-for record prices.

15
????????????? ???????? Penn Tree Bank

• LOC location NEG negation marker
• TMP time MOD modal verb
• MNR manner EXT extent, numerical role
• DIR direction PRP purpose
• CAU cause ADV general- purpose modifier

16
????????????? ???????? Penn Tree Bank

• Mary called John an idiot. (LABEL)
• Mary called John a cab.(SUMMON)
• Arg0 Mary Arg0 Mary
• Rel called Rel called
• Arg1 John (labeled) Arg2 John
  (benefactive)
• Arg3-PRD an idiot (attr) Arg1 a cab
  (thing summoned)

17
???????????? ???????? RST Bank

• RST Discourse Treebank Corpus 385 ??????? ??
  Wall Street Journal
• ( Root (span 1 156)
• ( Nucleus (span 1 129) (rel2par Topic-Drift)
• ( Nucleus (span 1 40) (rel2par Problem-Solution)
• ( Nucleus (span 1 21) (rel2par span)
• ( Nucleus (leaf 1) (rel2par span) (text _!Kidder,
  Peabody
• Co. is trying to struggle back.ltPgt_!) )
• ( Satellite (span 2 21) (rel2par circumstance)
• ( Nucleus (span 2 5) (rel2par Contrast)
• ( Nucleus (span 2 4) (rel2par span)

18
Penn Discourse Treebank discourse connectives

• (1) all subordinating conjunctions
• (2) all coordinating conjunctions
• (3) certain adverbials
• (4) implicit connectives between adjacent
  sentences.

19
Penn Discourse Treebank discourse connectives

• Subordinating conjunctions
• temporal (e.g., when, as soon as)
• causal e.g., because)
• concessive (e.g., although, even though)
• purpose (e.g., so that, in order that)
• conditional (e.g., if, unless).

20
Penn Discourse Treebank discourse connectives

• Coordinating conjunctions
• and, but, or.

21
Penn Discourse Treebank discourse connectives
(adverbials)

• (4) In the past, the socialist policies of the
  government strictly limited the size of new steel
  mills, petrochemical plants, car factories and
  other industrial concerns to conserve resources
  and restrict the profits businessmen could make.
  As a result, ()industry operated out of small,
  expensive, highly inefficient industrial units.
• (5) Strangely, (-) conventional wisdom inside the
  Beltway regards these transfer payments as
  uncontrollable or nondiscretionary.

22
???????? ????????????? ?????????

• EXP ID"54"gt ltREF SRC"53"/gt ltWgtWilliamlt/Wgt
  lt/EXPgt ltWgtandlt/Wgt ltEXP ID"55"gt ltREF SRC"53"/gt
  ltWgtMargielt/Wgt ltWgtHammacklt/Wgt ltEXP ID"56"gt ltREF
  SRC"55"/gtltWgtMrs.lt/Wgt ltWgtHammacklt/Wgtlt/EXPgt lt/EXPgt
  ltEXP ID"59"gt ltREF SRC"54 56"/gt ltWgtThelt/Wgt
  ltWgtHammackslt/Wgt ltWgtapos7lt/Wgt /EXPgtltWgtownlt/Wgt
  ltWgthomelt/Wgt